E-MailThe Principal-Agent Problem: An Old Alignment Issue with New Urgency
by John Berry, Fellow, Cutter Consortium
If frequent mention of business-IT alignment were a capital offense, I would have been executed long ago instead of writing this Executive Update, which intends to amend my various comments in earlier Cutter publications about business-IT alignment. It's no crime to champion the cause of an important business issue, but I must plead guilty to a misdemeanor in overlooking an important subtlety, a fine distinction in the topic whose importance is heightened in our current economic mess.
Alignment is frequently explored at an organizational level between IT and business units. Whether the context be investment decision making, decision-making governance, or even a specific management methodology such as the IT Infrastructure Library (ITIL), alignment defines that condition in which the IT organization and business units work collaboratively to deliver technology or services that support business objectives. Resources and brains are marshaled collectively within each department to deliver on that common objective. Fine, as far as it goes.
However, the management literature for years has discussed what is known as the principal-agent problem, a pathology arising in the behavior of individuals who work in these departments -- those with enough power to undermine business-IT alignment.
The crux of the principal-agent problem is this: the principal -- in this case, an organization -- and the agent chosen to represent it -- the employee -- sometimes work at cross-purposes, with perverse outcomes. Often, an agent has a personal agenda that competes with the best interests of the organization when his or her true fiduciary responsibility is to support what is in the best interest of the company. Competing agendas can manifest themselves in several ways, an explanation of which is the purpose of the rest of this Update. CIOs and IT managers would do well to pay attention for incidents of principal-agent bias in decision making around IT, especially now, when poor investment decision making in a very challenging business climate can magnify the financial problems that the organization overall faces.
Three principal-agent problems undermine company performance. They are explored in no order of detail, but all are framed within the context of alignment or, more accurately, misalignment. Here's what the CIO and his or her team should watch for.
MISALIGNMENT ARISING OUT OF CONFLICTS OF TIME
The organization's investment timetable might be very different from that of a hotshot manager who quickly moves from position to position in building broad-based managerial experience. GE has been known for exposing senior managers to many facets of its diverse businesses. However, in organizations where managers move frequently, they might have the tendency to favor projects, including IT investments, with a short payback. Why bother supporting an investment with a longer-term horizon whose payoff might emerge well past the time when the manager has moved on? That manager gets no credit for the success of the project. In a difficult economy, short-payback projects might seem particularly attractive because organizations are acutely focused on how quickly the organization can recoup its original investment, as opposed to riskier future value-creation possibilities from an investment à la ROI. Yet a bias toward short-term goals exclusively leaves longer-term opportunities unexplored and underinvested.
MISALIGNMENT ARISING OUT OF CHAMPION BIAS
In dysfunctional organizations, all sorts of decisions get made having nothing to do with objective decision making. Relationships between a project champion and senior managers in a position to see a project win approval becomes a deciding factor in moving the initiative forward. Projects rooted in this kind of ad hoc governance are successful because of luck; they succeed despite any empirical evidence of their prudence. Yet, even organizations embracing the rationality that derives from disciplined governance are vulnerable to champion bias in the following way.
Consider the electronic manufacturing company contemplating a multiyear investment in IT that would radically transform its complex product delivery environment. This company has many customers across the globe. The goal was to allow more local autonomy and control of inventory and shipments near the point of manufacture so that product could be delivered more quickly. A big software investment would allow such a reengineering of critical customer-oriented processes. Turnaround times to customers were costing the organization money and customer goodwill. It wanted to act fast. Given the scope of the project, the CEO became directly involved in the decision making.
On one side, the seasoned CFO questioned the forecasts of expected returns. While agreeing with the strategic intent of the investment, she felt the numbers were optimistic, in both their size and in how quickly the investment would generate value. Based on his understanding of the financial model and an understanding of the complexities to integrate a suite of sophisticated applications, the CIO concurred. On the other side, an operations manager enthusiastically pushed for immediate investment. As it happened, this manager's scope of responsibility and turf would expand significantly from an organizational reengineering arising out of the investment. This operations manager was also a trusted confidant of the CEO, who chose this individual and a few others to groom for top positions in the company.
In the end, the CEO sided with the operations manager. The result was an 18-month delay of the go-live date because of integration complexities and substantial business process reengineering that the system required. Several customers left for a competitor. While the CEO was open to all perspectives, the deciding factor in approving the project was the influence the project champion had on the CEO. This bias proved a value killer.
MISALIGNMENT FROM DIFFERING RISK TOLERANCES
In economic downturns as seemingly steep as this one, organizations are highly risk averse, yet some risks are worth taking; after all, returns from an investment usually run proportional to the size of the risk. Low-risk investments offer value, but the returns are relatively low compared to a riskier project that has a higher likelihood of failure but will deliver significant strategic impacts. As risk intolerant as organizations are in times of economic turbulence, managers are likely even more risk averse. An organization can profess that a bad investment decision will not kill a manager's career, but the fact is managers are terrified of making a poor decision they believe could undermine their career trajectory, especially when the investment is large, high profile, and made in an environment of economic uncertainty.
While avoiding the risk of reputation to the cause of a certain promising investment by not recommending it, the manager exposes the organization to the opportunity cost of not investing, the opportunity forgone. Yet to the manager, the consequences of commission -- recommending a risky investment that fails with the organization and his reputation suffering the impact -- are infinitely less attractive than the consequences of omission -- doing nothing, preserving his reputation, and having the organization suffer the opportunity cost of what could have been a huge value creating IT investment. "Omission bias" is a well-known phenomenon in management whose incidence can be heightened during this difficult business climate.
CONCLUSIONS
To understand how alignment can be achieved, it is not sufficient to explore it from an organizational perspective alone when influential employees can see to it that business objectives and IT are misaligned. Unfortunately, there are no easy solutions to the principal-agent dilemma.
Confronting Time Bias
In the case of the shortsighted manager, organizations can inoculate themselves from this bias by ensuring the input of senior managers in investment decision making for the simple fact that they are likely committed to the organization for the long haul. Their investment perspective is likely longer term than that of the ambitious manager ready to launch into his next gig. Payback is a very attractive metric in recessions because managers are finely attuned to the effectiveness and productivity of capital. Yet prudent managers also know that some IT investments must consider the longer-term impacts and competitiveness of the company, even if so much decision making today comes with short-term pressure.
Confronting Champion Bias
How to recognize champion bias? That occurs when one manager with an especially close relationship to senior decision makers aggressively supports an IT investment while the majority of managers, armed with hard data, are against it. The champion bias problem might be most effectively handled with a persuasive, point-by-point counterargument from the CIO and allies to the assertions made by the IT investment's chief evangelist. This tactic will not always win because often the champion bias pathology is introduced into investment deliberations precisely because managers suffer a lack of solid hard data. The best the CIO can hope for is the ability to recognize champion bias when the condition exists as well as the opportunity to present as persuasive a counterargument that can be mustered.
Confronting Risk Bias
Because it is human nature to magnify unpleasant events and discount pleasant ones, memories of investments gone bad are etched in granite. Who wants to work in an environment of long institutional memories when it happens to have been you that was responsible for one of those poor performing investments? Senior management must foster a culture of risk taking in smaller decisions so that the company is closer to risk neutral (a balanced view toward risk, not one indifferent to risk) when confronted with larger, higher-stakes decisions.
One potential approach to overcoming the misalignments arising out of differences in risk between the principal and agent is the adoption of a portfolio approach to IT projects. Pure portfolio management is rooted in finance in which various investments are viewed as a bundle of value-generating vehicles each possessing its own risk profile. A portfolio management perspective of IT is designed for a more holistic view of discrete IT projects, their financial profiles, and project status, among other data points. One argument holds that if managers could view each IT investment as one in a series of opportunities to create value rather than viewing each investment as an isolated event, then managers' risk tolerance would improve. The only problem with this argument is that although a series of IT investments might indeed have the attributes of a portfolio, each proposal is judged on its own merits. It is difficult to reconcile the suite-of-investment-opportunities theme with the necessary realities of investment decision making, which looks at value creation discretely, one investment at a time.
It is said that success has many fathers, but failure is an orphan. Actually, the principal-agent problem has many fathers, even though managers would prefer to leave the failures that this management pathology causes at the orphanage door. In the end, everyone knows who fathered the misaligned IT investment idea. The hope is that managers can recognize the principal-agent bias where it lurks and stop the monster before it compromises the organization's IT investment performance.